Chronic Exposure to Low Doses of Dioxin Promotes Liver Fibrosis Development in the C57BL/6J Diet-Induced Obesity Mouse Model

Background: Exposure to persistent organic pollutants (POPs) has been associated with the progression of chronic liver diseases, yet the contribution of POPs to the development of fibrosis in non-alcoholic fatty liver disease (NAFLD), a condition closely linked to obesity, remains poorly documented. Objectives: We investigated the effects of subchronic exposure to low doses of the POP 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), an aryl hydrocarbon receptor ligand, on NAFLD progression in diet-induced obese C57BL/6J mice. Methods: Male C57BL/6J mice were fed either a 10% low-fat (LFD) or a 45% high-fat (HFD) purified diet for 14 weeks and TCDD-exposed groups were injected once a week with 5 μg/kg TCDD or the vehicle for the last 6 weeks of the diet. Results: Liver histology and triglyceride levels showed that exposure of HFD fed mice to TCDD worsened hepatic steatosis, as compared to either HFD alone or LFD plus TCDD and the mRNA levels of key genes of hepatic lipid metabolism were strongly altered in co-treated mice. Further, increased liver collagen staining and serum transaminase levels showed that TCDD induced liver fibrosis in the HFD fed mice. TCDD in LFD fed mice increased the expression of several inflammation and fibrosis marker genes with no additional effect from a HFD. Conclusions: Exposure to TCDD amplifies the impairment of liver functions observed in mice fed an enriched fat diet as compared to a low fat diet. The results provide new evidence that environmental pollutants promote the development of liver fibrosis in obesity-related NAFLD in C57BL/6J mice. Citation: Duval C, Teixeira-Clerc F, Leblanc AF, Touch S, Emond C, Guerre-Millo M, Lotersztajn S, Barouki R, Aggerbeck M, Coumoul X. 2017. Chronic exposure to low doses of dioxin promotes liver fibrosis development in the C57BL/6J diet-induced obesity mouse model. Environ Health Perspect 125:428–436; http://dx.doi.org/10.1289/EHP316

. Body weight (BW) gain and plasma leptin concentrations after 5 weeks of LFD or HFD.  Figure S1. Distribution of potential low and high gainer individuals in the "TCDD" and "control" subgroups at week 5 of the LFD/HFD intervention. After mice received either a LFD or a HFD for 5 weeks (n=30), body weight (BW) gain and plasma leptin levels were determined, as described in the Methods, to randomize the mice into two subgroups for future injections (from week 8) with 5µg/kg of TCDD (LF-tcdd, n=16 and HF-tcdd, n=16, respectively) or the vehicle (LF-ctrl, n=14 and HF-ctrl, n=14, respectively). A. Variability in BW gain and plasma leptin concentrations at week 5 within the four different subgroups of injection. B. Potential low and high "gainer" status of the mice based on body weight gain and leptin level at week 5 (see also Table S1). In the HFD group, a BW gain of 5 g (and over) and a plasma leptin level of 5 ng/mL (and over) (red lines) were chosen to discriminate potential high gainers (zone 1) and low gainers (zone 2). In the LFD group, low and high gainers do not differentiate yet at week 5, based on those parameters (zone 2). However, the mice with barely detectable leptin levels or with a level of 2.5 ng/mL (and over) might be considered as potential low or high gainers, respectively. All the animals were divided equally in the control and TCDD subgroups. As mentioned in the Methods, two animals were excluded from the study at the time of sacrifice (animals 261 and 227), which belonged respectively to the LF-ctrl and LF-tcdd groups. Data are expressed as mean±SEM; a, versus LF-ctrl; b, versus LF-tcdd; p<0.001.   Tables   Table S1. Body weight (BW) gain and plasma leptin concentrations after 5 weeks of LFD or HFD.
* *: mice that were excluded from further analyses at the time of sacrifice. Potential high gainer (orange) or low gainer (green) individuals are indicated, based on the two parameters (BW gain, leptin) in the Table. See also Fig. S1. a, versus LF-ctrl; b, versus LF-tcdd; p<0.001.
After an acclimatization period of 2 weeks, the mice were injected intra-peritoneally (200 µL/25 g) once a week with either TCDD (0.1, 1, 10 or 25 µg/kg) (LGC Standards) diluted in corn oil or the vehicle (corn oil, Sigma), during 6 weeks. Five days after the last injection, the food was removed from 8 AM to 2 PM. The mice were then anesthetized with isoflurane and blood was drawn through retro-orbital sinus puncture prior to sacrifice by decapitation. The liver was removed, weighed, and either snap-frozen in liquid nitrogen or, for histology, fixed in buffered formalin and processed for paraffin embedding. Serum and plasma samples were obtained after centrifugation of the blood. All of the samples (ctrl n=4, TCDD0.1 n=5, TCDD1 n=5, TCDD10 n=4, TCDD25 n=3) were stored at -80°C until use.
Mice were housed and treated as described in the Methods.

Physiologically based pharmacokinetic modeling
In the present study, the rodent physiologically based pharmacokinetic model developed by Emond et al. in 2004 was used.

Determination of the threshold dose of TCDD that induces liver fibrosis
The preliminary objective of the study was to determine the threshold subchronic dose of TCDD which induces liver fibrosis. Male C57BL/6J mice were injected once a week with increasing doses of TCDD ranging from 0.1 to 25 µg/kg or its vehicle during 6 weeks and several hepatic parameters were measured. As a marker of exposure to TCDD, we showed that the induction of liver Cyp1a1 mRNA increased with increasing doses of TCDD ( Figure   S2A).
Although there was no significant alteration in body weight (BW) during the course of the experiment (data not shown), chronic exposure to increasing doses of TCDD gradually induced hepatomegaly in the mice. The liver weight (LW), expressed as a percentage of the body weight ( Figure S2B), increased significantly at a dose of 1 µg/kg of TCDD and represented more than 6% of body weight in mice injected with the highest doses of TCDD (10 and 25 µg/kg). Similarly, liver triglyceride content was moderately increased at 1 µg/kg TCDD (1.5-fold) with large increases at the two highest doses of TCDD tested (2.7 and 3fold, respectively, for the 10 and 25 µg/kg doses) ( Figure S2C). In addition, the 2 highest doses of TCDD led to an increase in serum alanine (Figure S2D) and aspartate (data not shown) aminotransferase activities, which indicates the presence of hepatic lesions.
Liver sections were stained with hematoxylin-eosin and picro-sirius red to evaluate the histopathological changes induced by TCDD. Consistent with the liver triglyceride content, the injection of increasing doses of TCDD for 6 weeks gradually induced liver steatosis. Mice which received an injection of 1µg/kg displayed negligible steatosis (less than 5% of the hepatocytes exhibited lipid droplets). Marked accumulation of lipids was observed in the mice injected with the highest doses of 10 and 25 µg/kg TCDD ( Figure S2E, red arrows). The 8 administration of the highest doses of TCDD (10 and 25 µg/kg) also was associated with an inflammatory cell infiltration ( Figure S2E, black arrowheads) whereas the mice injected with the lowest doses of TCDD (0.1 and 1 µg/kg) did not present any signs of liver inflammation.
More importantly, liver staining with picro-sirius red showed that the highest doses of TCDD also induced collagen fiber deposition ( Figure S2E, large black arrows) whereas the lowest doses had no effect.
To characterize, at the molecular level, the response to the injection of increasing amounts of TCDD, the levels of mRNA of several genes involved in lipid accumulation as well as those of genes that are markers of inflammation and fibrosis were measured by qPCR. All the genes exhibited dose-dependent effects although the threshold dose differed among the genes (Figure S2F,S2G and S2H). The level of Cd36 mRNA, a membrane transporter involved in lipid uptake and a direct AhR target gene, was increased in all of the mice injected with TCDD as compared to vehicle-treated mice. In contrast, for Pparg, a nuclear receptor regulating lipid storage, Plin2/Adrp, a major protein for lipid droplet formation, and Acta2/aSma, a typical myofibroblast marker, only the mice injected with 10 and 25 µg/kg TCDD exhibited increased levels of mRNA (Figure S2F and S2H) which is consistent with the histological observations ( Figure S2E). In contrast, the mRNA level of Dgat2, a protein involved in triglyceride synthesis, associated also with liver damage, significantly decreased with injections of increasing concentrations of TCDD, starting at 1 µg/kg ( Figure S2F). Furthermore, the expressions of the macrophage marker genes Cd68 ( Figure S2G) and Itgam/Cd11b (data not shown) as well as the major pro-fibrogenic cytokine Tgfb1 and the typical myofibroblast marker Col1a1 ( Figure S2H) increased with increasing amounts of TCDD injected starting at 1 µg/kg. Fibrotic scars were observed for all the mice that were injected with 10 and 25 µg/kg TCDD whereas the levels of mRNA of genes that are markers for inflammation and fibrosis were already mildly induced with 1 µg/kg TCDD, indicating that the threshold TCDD dose for fibrosis is encompassed between 1 and 10 µg/kg. Therefore, we chose the intermediate dose of 5 µg/kg TCDD for the diet experiments.
In parallel, in order to determine the relevance of a weekly injection of TCDD in mice over a period of 6 weeks to human exposure, we used a physiologically based pharmacokinetic model. With this model, the intra-peritoneal injection of 5 µg/kg TCDD is predicted to give a final concentration of TCDD in the serum of mice that is below 70 ppt wet weight, which is coherent with values for highly exposed human populations ( Figure S3) (Emond et al. 2005).  Figure S1. Distribution of potential low and high gainer individuals in the "TCDD" and "control" subgroups at week 5 of the LFD/HFD intervention. After mice received either a LFD or a HFD for 5 weeks (n=30), body weight (BW) gain and plasma leptin levels were determined, as described in the Methods, to randomize the mice into two subgroups for future injections (from week 8) with 5µg/kg of TCDD (LF-tcdd, n=16 and HF-tcdd, n=16, respectively) or the vehicle (LF-ctrl, n=14 and HF-ctrl, n=14, respectively). A.

Supplemental Figures
Variability in BW gain and plasma leptin concentrations at week 5 within the four different subgroups of injection. B. Potential low and high "gainer" status of the mice based on body weight gain and leptin level at week 5 (see also Table S1). In the HFD group, a BW gain of 5 g (and over) and a plasma leptin level of 5 ng/mL (and over) (red lines) were chosen to discriminate potential high gainers (zone 1) and low gainers (zone 2). In the LFD group, low and high gainers do not differentiate yet at week 5, based on those parameters (zone 2).
However, the mice with barely detectable leptin levels or with a level of 2.5 ng/mL (and over) might be considered as potential low or high gainers, respectively. All the animals were divided equally in the control and 11 TCDD subgroups. As mentioned in the Methods, two animals were excluded from the study at the time of sacrifice (animals 261 and 227), which belonged respectively to the LF-ctrl and LF-tcdd groups.